Hydronephrosis
| Article DOI | 10.3389/fped.2020.00001 |
| Objective | To predict SFU grade of hydronephrosis |
| AI Approach | CNN; Layer-wise propagation to visualize output. |
| Data Source(s) | Institutional series (2420 sagittal hydronephrosis ultrasound images) |
| Model Input | 256 x 256 pixel images of sagittal ultrasound |
| Model Outcome | SFU 0-4 Mild vs. Severe hydronephrosis SFU 2 vs. SFU 3 |
| Model Metrics | SFU 0-4: F1 = 0.49, accuracy = 51% Mild vs. Severe hydronephrosis: F1 = 0.78, accuracy = 78% SFU 2 vs. SFU 3: F1 = 0.71, accuracy = 71% |
| Model Usability | NA |
AIPEDURO 
Leave a comment